Research On Travel Route Recommendation Based On Attention Mechanism And User Profile

Most tourists are constrained and influenced by multiple factors when formulating travel plans.The insufficient consideration of tourists’ personalized travel preferences and constraints in existing research makes it hard to meet tourists’ personalized tourism demands.From the perspective of user demand,this paper takes the tourism route recommendation system as the research object,introduces the user demand theory to divide the dimensions of the multi-dimensional profile model,studies the attraction recommendation method based on attention mechanism and profile,and constructs a tourism route planning method based on multiple constraints according to the tourism cost factor in the actual tourism scene,which provides a new idea for the optimization of personalized travel route recommendation service.This paper conducts relevant research based on real data sets from tourist attraction recommendations and travel route planning.The main contributions of this research are as follows:(1)The demands of users in different situations will be diverse.This paper introduces context factors to enrich the user profile dimension and provide users with targeted and contextualized travel recommendation services.At the same time,based on the theoretical basis and construction method of user portrait,the concept of "attraction profile" is innovatively proposed to strengthen the connection between users and attractions and improve recommendation efficiency.(2)Aiming at the problem that the current tourism recommendation system relies on user ratings and lacks research on other types of interactive behaviors,firstly,this paper uses a multidimensional profile model to describe users and attraction resources in an all-round and multi-level,to accurately reflect the characteristics of users and attractions,mines a large number of tourism texts to extract text features at the same time.Then,this paper uses the attention mechanism to assign attention weight coefficients to interactive behavior features and text features and finally proposes an attraction recommendation algorithm based on the attention mechanism and profile.Finally,this paper makes a detailed experimental comparison with other mainstream recommendation algorithms.(3)This paper systematically analyzes the problem of tourism route planning,reasonably sets the model assumption,and proposes a multi-constraint tourism route planning method for actual tourism application scenarios.This method is improved based on the nearest neighbor strategy of the greedy algorithm to calculate the optimal route cluster,including several tourism routes that meet users’ personalized travel preferences and constraints,then adjust the recommendation results according to the congestion coefficient of the road section,to maximize tourism revenue and improve tourists’ tourism experience.